Quality

[neat shard]

I dunno if it's be 80-90%... maybe like 30% at most

[north yew]

are you basing that on anything or is it just vibes

[neat shard]

Are you basing 80-90% on anything or is that just vibes?

[north yew]

I've mathcrafted several specific scenarios and actively looked for a recipe in 1.1 that doesn't have at least one short-chain ingredient bottlenecking its quality ratio

didn't find one

we can follow it through: gets bottlenecked by having a short chain, then and T1 modules get bottlenecked by the ratio, then T2 modules get bottlenecked by the ratio, then T3 modules again get bottlenecked by the ratio.

so ultimately you're getting one effective quality step for having quality modules in your assemblers

[neat shard]

Hmm, I might try looking into it myself

[north yew]

whereas if you used prod in all of those module slots instead, you'd be getting a guaranteed, multiplicatively-stacking bonus to efficiency at every step

[neat shard]

Perhaps you could work around the bottlenecks by disabling the qual modded assemblers when there's a stockpile of quality outputs, so the inputs can get fed into prod modded assemblers instead

[north yew]

sure, but in that case the prod modded ones would be doing the bulk of the processing anyway. very little juice for a whole lot of squeeze

[neat shard]

squeeze?

[north yew]

play on the phrase "the juice isn't worth the squeeze"

[open halo]

oh man I love squeeze

[neat shard]

Getting high-quality modules without getting garbage you then have to throw in the recycler might be worth it, just because of how many resources you waste by having to recycle stuff

[north yew]

even with in every slot, the bulk of outputs are gonna be normal and uncommon anyway. if you only care about best quality stuff, it's extremely expensive either way

the main thing is that if prod didn't exist things would be different, but with intermediates every single slot taken up by a quality module is a slot not taken up by prod

productivity in a chain is a steep exponential curve

with each step occupied by prod, you're guaranteed to take a step up to a steeper part of that curve. with quality, it's not even close

[jaunty citrus]

Basically when you have different length production chains for different ingredients, putting in everything has two impacts.

1. It makes it so that different quality levels have different ratios of ingredients, which doesn't work for equally consuming all products of all quality values.
2. You loose out on the opportunity cost of productivity, which has larger and more consistent benefits.

The only place you can reasonable use quality modules are then:

1. In the first step of each ingredients production chain. You'd then need separate assemblers for each quality level, which would increase complexity but may be worth it. The best place to do this would theoretically be in miners, as the opportunity cost of not using prod modules is lessened greatly by mining prod.
2. In the last step of a production chain (say the assembler that actually makes the T3 module, personal equipment, etc.
3. Inside recycling loops that are a part of a separated quality factory

The limitations applied by location 1) for quality modules will, in my opinion, make it an extremely unpopular choice for people not really enthusiastic about getting into quality production.

[tall sandal]

Quality loop plastic until you have the correct ratio

[rare thistle]

the correct ratio is all

[north yew]

it's definitely funny to me that people who just take the naive approach of throwing quality modules in mall products and then maybe eventually setting up recycle loops for especially important things are going to btfo all of the elaborate theorycrafted approaches that people are cooking up

[sick dock]

i still think early game, like when you're first getting off it might be good to do quality & . you won't have the recycler, and will probably want if you can get your hands on them

there's a space for it

[jaunty citrus]

Early on, it'll probably be pretty normal to get some quality modules into your production, so you can skim off the quality ones and use them to 'manually' craft some stuff

[sick dock]

you can essentially "recycle" Q1 intermediates by thworing them back into your production

so your buffers don't fill up excessively

[jaunty citrus]

getting +60% on your productivity for the rocket silo will be very good

[sick dock]

apparently you only need like...5 rocket parts now? that's so low when you just say it like that

it's still good though, as , are still quite expensive

[tall sandal]

this isnt actually insane. say you want . you can put quality in everything until you get to the assemblers. Then, you quality loop the and until its all . Then you send that though the prodded assemblers. How bad is would that be?

[jaunty citrus]

Yes, but you'll also need, most likely, dozens of rockets before even getting to your first planet other than

[sick dock]

productivity is freaking busted, isn't it

[rare thistle]

what's the math on prod? they're worse than aren't they?

[sick dock]

true. your rocket's vanilla purporse of 1k replaced with carting around all your fun toys

[north yew]

looping in the middle is exactly as wasteful as looping at the end

[sick dock]

4% * 1.6 = 6.4%, i believe

so yes, but it's still strong

are crazy expensive, even at Q1

[rare thistle]

oh that's even worse than I thought. eep

[jaunty citrus]

Except you'd most likely need more machines to loop at the same rate, so you'd need more high quality modules

[tall sandal]

no, because the ingredients have gone though several quality steps. and there's no ratio issue

[north yew]

are 6% so you would get 9.6% from

[sick dock]

remember. a isn't like, a dead-end product. you can always put it into a later, or simply recycle it for parts

[rare thistle]

wait 4% is its 6% for beat me to it

[sick dock]

ah, ty. i was like i don't think this is right

[jaunty citrus]

@tall sandal wrong reply?

[sick dock]

yeah, sry, got confused

maths correct, inputs wrong

[oblique wharf]

wow

[tall sandal]

you need a minimum of 7 rocket launches to even unlock space platform thrusters, and an 8th to board the space platform. that doesnt include the number of launches needed to actually build the platform itself

[sick dock]

i almost never use so i forget their #'s

[jaunty citrus]

Yes, you'd need probably dozens of rockets to get to your forst planet

[rare thistle]

I always have to work backwards from the 3rd tier and am never confident I did it right

[sick dock]

so actually is almost just as good at

noteworthy

[tall sandal]

oh and a 9th rocket launch to launch up enriched uranium too

[sick dock]

i think the cargo will only be a satelite? so you can't cart stuff up

[north yew]

within a specific step the odds of upgrading and the amount lost doesn't change whether you do it at the start or at the end. also this approach requires a lot more machines and comes with a much larger cost

[jaunty citrus]

And rockets for all the stuff you want to bring to the first planet

[sick dock]

but remeber, it's 20x as cheap, so if you send 10 rockets thats already half of vanilla

and we got quality, although i believe is unlocked on another planet, probably with its productivity theme

[tall sandal]

I wonder how much space platform you can send up per launch

[sick dock]

they'll have to work out the design space for that 😎

[final nimbus]

Rockets end up costing far less than 5% of vanilla

[rare thistle]

I bet they want us to have to scale up before we make it off planet

unless we do that improved ammo oneway ticket strat

[jaunty citrus]

https://cdn.factorio.com/assets/blog-sync/fff-382-rocket-silo-gui.png

I imagine a space platform to cost between 5-10 rockets with these numbers?

depends on your design of course

[tall sandal]

hopefully the majority of that weight is from blue belts

[sick dock]

i still cannot get over is 25% productivity, that is just insane. 100% with 4 of the buggers, a whopping 125% + base 50% in EMPs. Your EMPS will make any product cost like 30% of its original cost, including other EMPs and modules

[rare thistle]

☝️ that's the heavy one there, yeah :P

[jaunty citrus]

I imagine the undergrounds to be quite heavy

[north yew]

after three steps using full prod em plants you've reduced the cost to ~5% of what it would take with regular assemblers

[jaunty citrus]

1.1 blue chips fuming

[sick dock]

We were having a...lively discussion if green belts & foundries can only be crafted on , and it seems the answer is yes. so i hope those undergrounds aren't too heavy, we have a lot of rockets & platforms to cart around

[north yew]

random thought connected to that: sulfuric acid to steam is exclusive to , but lava processing isn't

[jaunty citrus]

I see no other reasonable interpretation for "exclusive recipes" other than limiting crafting location to that planet

[sick dock]

yes, and while I doubt that non-intermediates will get productivity researches, almost certainly and maybe will, reducing the cost even even further

[rare thistle]

yeah they wouldn't do that to us. belts'll be light (I say manifesting it as hard as I possibly can)

[sick dock]

Fulgora will be pumping out modules like no tomorrow, not to mention the ones you get "for free" from Scrap

[jaunty citrus]

Nah each planet will have to do it's own module IMO, so fulgora will have , vulcanus , etc.

[sick dock]

minecraft lava bucket here we come

i mean if the game will let you barrel molten iron in a freaking metal barrel then i feel we can do the same with lava

[neat shard]

@north yew Here's an example for you to chew on. Say you want one single Q2 , and you have unlimited access to Q1 s and Q1 s for your assembler 3s.

If your approach is to make it by throwing Q1 circuits into a prod 4 assembler with four qual mods, then recycling the components in an unmoduled recycler, it'll take on average 7.5x the base cost of a Q1 to get your desired Q2 . (Each iteration costs 0.75x, and on average it takes 10 iterations to strike paydirt)

Now suppose your approach is to make green and red circuits in qual modded assemblers from Q1 components, so as to stockpile Q2 components with which to directly craft a Q2 . Your assemblers print out 10x worth of circuits, of which 9x is Q1 and 1x is Q2. (Assume the Q1 circuits get fed into science or something so they're not wasted.) You missed out on 4x worth of circuits by replacing the prod mods with qual mods, so it ultimately took you 5x the base cost of a to get a Q2 – cheaper than the recycler loop approach.

[jaunty citrus]

I'm 90% sure we won't be able to do that

[sick dock]

we might have to make tungsten barrels or something for molten materials

tungsten pipe? prob not

[jaunty citrus]

No alternative pipes for now (they were too annoying in playtesting)

[neat shard]

Not every fluid can be barreled. For example, steam can't be barreled

[north yew]

Since we have recycling, it would be more efficient to transfer some resource-dense items, like express underground belts, to be recycled into the iron at the destination, compared to just sending the iron directly.
a driving force of the weight system is to make items that are very recyclable unenticing to use as a workaround

[neat shard]

So it's not unreasonable to suspect lava won't be barrel-able either

[sick dock]

melting point 1,000 C*, melting point ~1,500; tungsten ~3,400

[rare thistle]

true but the non-prodable items have much more leeway

[north yew]

(belts are now proddable)

[rare thistle]

beyond the 50%?

[sick dock]

also, minor gripe, but can they simply condense all the "fill fluid barrel" & "empty fluid barrel" into a single UI icon?

[jaunty citrus]

Technically productivity modules are still incompatible with the recipe

[sick dock]

it's a ton of UI space for literallly like...no meaningful benefit then other to see the fluid types

[jaunty citrus]

but the in-built prod from machines still applies

[neat shard]

That would be more complicated than it sounds at first

[jaunty citrus]

If you ask me, Barrels are a legacy feature from before fluid wagons existed, and should be removed.

[sick dock]

they are nice to store for far away locales; i've had to do that with outposts a few times

[rare thistle]

as long as its not more efficient to recycle belts than to ship the same weight of its components then I don't think they're going to be too strict with the weights

[sick dock]

are the real legacy feature

[rare thistle]

they can adjust it for gameplay feel

[north yew]

this is correct, but:
A) it only works in narrow, carefully-controlled circumstances like this
B) it's only considering the one best-possible-quality output
if you consider how many modules total get made for the resources, prod wins out. even if you multiply the uncommon ones by 1.3 and the rare ones by 1.6 etc etc

[jaunty citrus]

There are other solutions that could be added for barrel's common use cases, but I know they won't remove them because of how entrenched they are

[sick dock]

as yousay that, they actually mentioend recycling undergrounds to potentially be a problem, as it could be more effiicent then simply shipping iron

idk how many people would ship but it does feel weird that this gimmick is better

[neat shard]

Barrels have their uses. For example, coal liquefaction takes a little heavy oil to kickstart, and it's most convenient to supply it by barrel

[jaunty citrus]

You could add another item, similar to ice, that is a direct conversion between heavy oil - (something)

[north yew]

I fill my perimeter supply train by botting barrels of light oil over to it from the section and I am not ashamed

[sick dock]

if only it were plausible to pickup storage tank and have it retain resources

but maybe too hard on CPUs, i wouldn't know

[neat shard]

Explain in more detail?

[north yew]

a) it only works out this way when you're doing last step - 1, which becomes harder and harder to do the more things you want quality on. following it through all the way is basically having a segregated raw ore -> product section for every individual product
b) if we treat a normal output as 1, an uncommon output as 1.3, a rare output as 1.6, and so on and so forth, prod in all intermediates results in a larger total for the same quantity of input resources. quality in intermediates only starts to enter the picture if you're at a point where everything but the best possible quality is completely irrelevant

[jaunty citrus]

Basically you're not wrong, it's a really good early-on setup for bootstrapping your way into some and fewer modules.
But the setup you describe isn't very scalable, and you get a fixed ratio of 9x and 1x which won't remain useful as you progress and want higher quality things.

A lot of the time you'll probably find it cheaper, quicker, and easier to build a proper hub and stick good quality modules into your hub assemblers

[sick dock]

important question: when you recycle a , will you always get or greater components?

[jaunty citrus]

Getting some or might be worth it for that, imo

[north yew]

yes it does not downgrade

[sick dock]

pretty sure i knew taht already, but good to hear it confirmed

gathering at least a few buffer chests of intermediates in the early game does have the benefit of letting you decide what items you want to build later on. if you're indecisive, or don't know what you'll need, having the components on hand might save you some time later

more of a convenience, but there's nothing wrong with that

[north yew]

to elaborate a little further on b:
mostly I just feel like it's going to be a bit of a noob trap to only think about the best things you can get. normal products are still useful and good. a normal quality is still useful, and you can get to your 4x for the rocket silo simply by dispersing the normal quality ones along your factory

strategically I just feel like it's gonna be a lot better to not care too much about getting the best possible outputs and focus on deciding where the unusually good outputs are going to do the most work

another example: you can psuedo-recycle by just putting the normal quality ones down on and that can get you a nice stash of higher quality ones to set up space platforms and bootstrap

[neat shard]

Fair enough

Though, all this does hinge on qual mods having to compete with prod mods (i.e. the best modules in the game)

[north yew]

yeah exactly

[neat shard]

Seems plausible to me that any recipe that refuses prod mods will get stuffed to the gills with qual mods instead

[north yew]

if prod mods didn't exist then absolutely go nuts with jamming those quality modules everywhere

[north yew]

agreed, it would be foolish not to

[jaunty citrus]

It would depend on the use case, but broadly any end products that don't accept productivity should get quality instead

[sick dock]

here we come

[north yew]

unless certain external demands ask for

are really gonna get their chance to shine in SA imho

[sick dock]

or you just don't care and simply want

[north yew]

nah speed modules on their own are not good

just build two things

[sick dock]

bloody well hope so. they're almost useless in vanilla

[jaunty citrus]

Bad take, reducing pollution is quite helpful

[sick dock]

if ya suck

[north yew]

saves you resources like just in a more indirect way. less pollution = fewer biters throwing themselves at your walls and less evolution

[neat shard]

I gesture in the general direction of SE, where eff mods reduce energy consumption by 40% at tier 1 to a whopping 1000% at tier 9... and they still suck

[north yew]

used tactically can drastically reduce the amount of time and resources you need to spend on protecting your base

[tall sandal]

how does 1000% energy reduction even work?

[north yew]

presumably there's other modules that can increase cost high enough to justify that

[neat shard]

Lower limit of 20% energy consumption still applies, but it offsets a ton of energy increases from other module types

[tall sandal]

oh. right!

[jaunty citrus]

It counteracts the +3000% energy consumption from the speed modules

or whatever silly number it is

[rare thistle]

just gonna drop this here. probably mods change it and maybe 2.0 will too but

[north yew]

on power is free real estate, but it seems like a big part of SA is making power a lot less free

[sick dock]

it is additive. increase poewr by a bunch, and will negate it additively

so a with >100% isn't like...what you might htink it is

[tall sandal]

I use eff modules when I'm doing no spoon runs a lot

[rare thistle]

sorry wow I'm slow today

[jaunty citrus]

get me my -125%

[north yew]

slapping them in miners ASAP can reduce the size of your pollution cloud to like a third of what it was before. way fewer biter problems

[sick dock]

i do hope they buff at least a little. the & in partiuclar seem like abysmal rates

[jaunty citrus]

put it in my for +62.5% speed and -22.5% energy consumption for machines around the beacon
(If I remember my math right)

[sick dock]

it's not untrue, but you can always just wall or blow up the biters

[north yew]

that costs time and resources, though. many of those resources are on consumable things

whereas you make a and it's just saving you power/pollution forever

[sick dock]

shrug

i got this next to my name for a reason, i am biased

[tall sandal]

eff modules are probably gonna be good on space platforms

[north yew]

probably the default

[tall sandal]

but I agree that eff2 and 3 seems pretty bad

[north yew]

like you have to have a good reason not to use efficiency

[sick dock]

quality ice, let's go

[tall sandal]

lol

[north yew]

is worth imho

[neat shard]

I haven't really read the FFFs about space platforms, but I know that in SE power is even less of a concern in space than on the ground

[jaunty citrus]

You should read the fffs then

[sick dock]

yes, but outside it they feel limiting in vanilla

[neat shard]

Plot twist, I'm illiterate

[jaunty citrus]

Because that couldn't be further from the case in Space Age

[north yew]

also

[tall sandal]

lol do they take a lot of power?

[north yew]

mostly just stinky

[sick dock]

centrifuges are fairly power hungry per unit

but you odn't need many, generally

[tall sandal]

that uranium dust

[north yew]

with in the drills and in the centrifuges nuclear power approaches being pollution-free

[sick dock]

might also see real use on

we'll have to see about that tho

[north yew]

too tbh

[sick dock]

if Foundries really are that hungry, we might consider it, even if we have to beacon them, lol

[north yew]

400% solar is 400% solar but space and geometry are very constraining, plus yeah beeg drill and foundry are both described as being power hungry

[sick dock]

foundaries sound interesting, tbh

[jaunty citrus]

Eff modules in beacons won't be good imo without using quality

[neat shard]

Just because it's hard to link electric networks across the mud?

[sick dock]

solar power purportedly drains the further yo uget from the Sun

[sick dock]

it is impossible to put on the oily sands

you need the landfill tech, whatever that is

if you get lucky and your islands naturally close together, you can appropriate a bunch of space to jsut be power generation

[jaunty citrus]

Solar power will be really good on vulcanus. 400% production and a 90 second day/night cycle is nutty

[sick dock]

but if lightning power is incredibly good then you won't even care

[sick dock]

prob rip to geothermal energy dreams then 😦

[jaunty citrus]

I don't expect the geometry to make it that difficult either, you can absolutely get creative putting panels around rail, in between builds, at outposts, etc.

[neat shard]

There's also the possibility of ferrying energy around as steam in fluid wagons if lightning power isn't enough

[north yew]

why and same power cost y u do dis wube

[sick dock]

wat we feedign to make the steam? from scrap, i suppose

also you need a lot of steam/s to keep engines running.

it's like water, which people normally don't think about because you just pump it

but when you have to train water, it can be super annoying

[rare thistle]

it's problem because you effectively have a power limit per island. so if miners are eating 60% of the power on that island you can get away with as big a train station as you want but might not have the electricity to do the processing right there

[neat shard]

Offshore pump heavy oil out of the mud, then convert the heavy oil into solid fuel

[north yew]

really the bottleneck there is water for the boilers

[heavy ember]

Which is why people tend to build nuclear where the water is

[neat shard]

Fair point

[sick dock]

oh, true

and yes, water is scarce, and you'd prob rather have that for cracking

[north yew]

considering that supercapacitors take water's definitely gonna be super precious

[rare thistle]

revisiting an idea: dropping extra ice from space?

[sick dock]

i had the same idea

i think you can't see asteroids in orbit, but you can make a simple boomerrang trip with a platform

[heavy ember]

Atlest with SE, water is precious on no water surfaces.

Even with how much water ice compresses for logistics bots

[sick dock]

and just dump off the 🧊

[neat shard]

I'm guessing it's not going to be that easy to get enough water to sustain a base

[north yew]

the main thing is that you only have one drop platform and all of your planet's stuff has to go through it

[neat shard]

If imports from space are too easy, the challenge of adapting your base to local conditions is moot

This is a fairly big issue in SE

[rare thistle]

fulgora's gonna require a sophisticated train network anyway, I'm not too worried

[heavy ember]

Even if it isn't, if it's possible people will do it.

I've seen rockets full of water barrels destined to a waterless planets in se

[rare thistle]

good point

[sick dock]

as i pondered it, rockets aren't the real limiting factor in logistics, esp with in your silos. getting them across space in your platforms might be the real bottleneck

you can ofc just make more platforms, but its the same as saying build more trains

i'm sure interplanetary logistics being well crafted is literally at the top of their design-philosophy document, so i won't worry about it too much

[north yew]

the main thing w/ rockets is that it adds a tax to anything you ship between planets

initially it's a very hefty tax, eventually it becomes less hefty, but there's still always that tax

[sick dock]

yes, but they're really trying to get you to use these 50% prod bonus buildings

[neat shard]

Is it? "Build more trains" runs into the problem where the trains get in each other's way, such that eventually adding more trains doesn't increase throughput any more

Do space platforms have a similar dynamic?

[sick dock]

who knows, but no, i don't think they can bump into each other

although i am still wondering: if a platform has its thrusters destroyed...how do you recover it?

can't exactly walk over to the area as if you would an unfueled train

[north yew]

honestly there's probably just an "abandon platform" button

as far as we know there's no way to pull up to one platform while riding another

[sick dock]

fair enough, but if you had 1k sciecne in there or w/e you're gonna be real bloody sad

better make sure your turrets hold up, then

but there will be growing pains for sure, for many players

[north yew]

feels like that's part of the design. don't move platforms with expensive things through dangerous routes unless you're sure they can handle it

[sick dock]

smart way to play, anyway

[north yew]

from the FFF's it seems like they designed it in a way to gently ease you in to it

[jaunty citrus]

Any platform that gets stuck in transit will most likely be destroyed, something on it was already destroyed for it to stop moving

[north yew]

first you do satellites then you build a basic platform that isn't under major duress then you have that platform start orbiting to harvest space rocks

[sick dock]

it also seemed like asteroids could only approcah the front 180* of a platform

so your thrusters - presumably in the back - would likely be the safest

not the most realistic for space, but obv this is a game first

[north yew]

what happens if you build thrusters facing opposite directions?

[sick dock]

https://tenor.com/view/you-spin-me-round-dead-or-alive-pete-burns-like-a-record-baby-right-round-round-round-gif-17323319

[north yew]

we simply crash the death frisbee in to the biter nest

step aside trainsaw a new biter destroyer is coming along

[neat shard]

Forgive me for making SE comparisons again, but in SE, losing the spaceship is usually more painful than losing the payload contained in the spaceship, because spaceships ain't cheap

[sick dock]

space platform is prob like...some iron & steel. no idea what the....ugh..."space platform starter pack" will cost

"initializer" would prob be more fitting. what is it with Wube and these weird names in SA

actually, space platform prob uses now i think about it

[jaunty citrus]

'Space platform starter pack' is probably just being descriptive, my bets on it just being called 'space platform'

[sick dock]

Birth of a platform
You start by launching an item called "Space platform starter pack" into space. The item is inserted into the rocket silo in a similar way to the satellite.

idk how to quote an FFF directly like others do, but tha'ts a direct quote

[north yew]

> like this

(don't forget the space)

[sick dock]

ty

"space platform" refers to the flooring of the platform

the actual, like "platform" part of the platform

[jaunty citrus]

I misremembered it, but yeah It's pretty self explanitory then so I don't really see an issue with the name

[weak vault]

You can have the crafting be productivity and the recycling be quality or both quality. Both reduce the cost of getting more and higher qualities. I wonder how do both choices progress over time and which one will be the best, will we have one choice for early game, another for mid and going back to the other for late game?

Considering that productivity is usually better, we can easily have 90% prod for recipes on buildings with 50% base + 4 . For late game i imagine some recipes will be able to reach the max productivity of 300% by utilizing the recipe research. This makes quality almost free in late game, except energy costs. However you can achieve some compound effects with quality of quality which could be the breaking point in mid-game.

[little orchid]

The thing is... There has never been an arbitrary module restriction on a building before (not counting beacons, they're a different beast). Recyclers can't accept prod, because it would create an infinite loop. Fluid recipes block quality modules, because quality fluids don't exist. Such a restriction is only applied when it directly clashes with game mechanics. I simply don't see why such an extreme special case should be made for this one building

And your reasons are iffy at best. The assembler 3 will still have plenty of roles, just not for circuit making. Why would the devs give us this exciting new way to make circuits, only for it to be "only sometimes better than the old way"

[unborn flax]

Oh absolutely there's some iffy stuff in there as we all have the same set of limited hard facts. The graph I posted isn't speculation though - it's the current reality. Feel free to ignore the green/cyan lines if you want.

Re: the assembler3, I don't mean "assembler3 would otherwise be unloved". Perhaps I phrased it poorly. I mean "it would be cute design that the optimal quality setup for EMP would involve both EMP and assembler3, without one building completely taking over the other".

Actually don't ignore the green / cyan lines - even if I'm wrong about the EMP, they're relevant for when prod modules are hypothetically allowed by the EMP but prohibited by a recipe (e.g. electric poles and similar non-intermediates)

[jaunty citrus]

Limiting prod modules on machines like that doesn't make a lot of sense when they give some recipes infinite productivity research, which would still work on those machines

[remote scaffold]

[sick dock]

My birthday is late April. Would love a reveal on my birthday wekeend 🙂

[iron root]

also: what makes the foundry 'prod themed' and the EMP ' quality themed'?

they both give 50% prod

in fact, what even makes the planets themselves themed towards prod or quality?

[sacred totem]

well, quality is obvious

fulgora is where you unlock quality

as for vulcanus and prod, eh, idk

[iron root]

its where you get quality 3s, but thats about it

I was initially on board with planets being prod/eff/speed/whatever focused but with the EMP having 50% prod as well, I reevaluated that and the more I think about it the less I think it makes sense

[neat shard]

"How hard is it to beat Space Age (and make infinite research) with only qual mods and no prod mods?" ~Video I would watch, or even be tempted to make myself

[north yew]

when it was just revealed it was red building + building with innate prod + the thing the building is doing being very prod-themed (taking your resources and stretching them much further)

has thrown all of that out the window, though.

[iron root]

and I've said it before but I think the devs did a good job recognizing just how strong productivity was and realized it would instantly outshine everything else if the other options didnt have prod too
So we're not getting focuses on prod/speed/eff, we're getting focuses on different sections of the crafting chain

even if the EMP doesnt get some of the extra-efficient recipes like the foundry does, the fact that it applies its innate prod to recipes you otherwise couldnt prod mod has a similar effect

[north yew]

the big thing is that the emp chain is much longer

w/ foundry the most you're getting is three in a row, and two for most things

but the em plant makes then it makes then it makes then it makes then it makes then it makes

[iron root]

yeah some of that gets really silly

[north yew]

foundry closes the gap w/ more efficient recipes, which are going to be more insane than most people think tbh

we saw today that takes 10 molten iron, and in most things in Factorio there's a pretty strong 10 fluid to 1 solid ratio

so likely that means is also 10

[iron root]

I spitballed some numbers based on what we were shown in the fff and that gear recipe puts out something like 10-12 gears per completion

2 4-stacked green belts almost full I said was about 400 gears/s, divide by 26 (for +2500% speed) and 1.5 (for foundry prod) and 10.25

and 400 gears/s is a lowball i think

[north yew]

390/s I'm pretty sure

did a bunch of math earlier and pretty concretely landed on a crafting speed of 5 for the foundry

[iron root]

ah yeah forgot the foundry itself probably has a cracked base crafting speed

at base crafting speed of 5 thats still 2 or 3 gears for 10 molten iron

assuming it follows the usual rules of buffering 2x its recipe cost

[north yew]

I arrived at 390 assuming it's 1x per 10 molten iron

[iron root]

we dont know the base recipe speed for this craft though

[north yew]

watching the inserters on the LDS foundry in Swimming in Lava shows it crafting about once per four seconds, so that lines up with the crafting speed of 5

[iron root]

thats a good catch, cause that was the default lds recipe iirc

[north yew]

has to be

[iron root]

so if we have a crafting speed of 5, and we know how many crafts/t we make at 26x speed
we can find the base recipe time

[north yew]

same as regular

1s recipe time would mean a crafting speed of 10

[iron root]

yeah nvm my math was bad

[north yew]

I was stumbling all over myself this morning, working backwards toward a crafting time from other variables is not something we need to do often

[iron root]

it is 260 crafts/s, and somehow I got turned around with that and the 26x speed

so yeah with that in mind 1 gear for 10 molten iron does sound right

[north yew]

likely the recipe for melting the iron is 1 ore to 10 output (before prod is factored in)

so basically there's a sneaky +100% prod working its way in when you go straight to gears, totaling +250% just from the foundry's innate properties

[iron root]

yeah 1:1 is nice and clean and would compensate for not abusing prod chaining melt->plate->gear
at least until higher level modules and techs kick in

[north yew]

with best quality prod modules that means you're getting 12.5x as many iron gears from the same amount of ore

vs just regular smelting then crafting recipe

[iron root]

actually, I wonder now if the foundry can even be used to go iron plate->iron gear, I don't remember seeing any confirmation this was possible?

[north yew]

nope

you either go direct to plates or direct to gear/steel/cable

[iron root]

I guess an assembler gets a max of 100% prod with , which is exactly as much as you'd get from direct-casting
and we won't have recipe prod for everything

[north yew]

yeah exactly I'm assuming that the efficiency of the direct recipes is based around your fancy new building not getting obsoleted eventually by the old way of doing things

[iron root]

so thats how they can keep the foundry in play instead of prod chaining blowing it away again

[north yew]

with it's better to direct smelt in the foundry in all cases if the and recipes take the same amount
with the best you're ever gonna get is +50%, so that recipe might be a little less efficient (especially b/c it's confirmed getting a prod research)
with it gets tricker. honestly it would be fine if foundry plates -> em plant cable was slightly more efficient than direct to cable from foundry, but it also still might be the design intent for the foundry to be the most efficient way.

[neat shard]

You know, I'm starting to get the impression that prod mods are so powerful they significantly warp Factorio's fundamental game balance

If multi-step prod mod chains are so powerful that even an efficiency bonus isn't enough to convince people to use anything but the recipe chain with the most proddable steps

Not to mention what we were talking about earlier, how the newly-introduced fun and cool quality module will pretty much always have to take a backseat to prod mods in any recipe that accepts prod mods

[north yew]

in the case where the efficiency bonus wasn't enough to justify skipping a step, that would just mean Wube did an oopsie with the numbers. it's specifically a case where the thing you get from prod - better resource efficiency - is the thing you would be getting from the alternate recipe

w/ the ratio problem it's actually good game design that the simpler, less headache-inducing approach is superior imho

quality is interesting, but building the same supply chain five separate times is less interesting

I do think it's correct to say that prod is extremely powerful and bends the game around it, but Wube are clearly conscious of that and do so in interesting ways. you make a Faustian bargian with the prod modules: your machines become very stinky, they become huge power hogs, and it becomes much harder to gauge exactly how many assemblers for a specific thing you need

(side rant: this is also why I consider rate calculators to be light cheating. it seems like a very intentional design choice that, without prod modules, ratios are very clean and easy to calculate through mental math alone. prod modules make it much harder to do that math, and it seems like that's on purpose imho)

(of course the very powerful tool is very powerful when you use an external tool or mod to completely sidestep one of its major downsides)

[iron root]

quality is giga funky in that you can continue to reap the benefits of quality after removing the module from the machine

other modules dont have the lasting impact quality modules do, which makes them sort of apples to oranges

but the fact you can duplicate quality materials with prod mods is another twist. It'll be a fun system to work with

[median verge]

Feels necessary for stuff like pY... but thats another discussion

[high zenith]

Has anyone done the math on whether mixing speed and quality modules is ever a good idea?

I'm thinking orange dot speed 1s and grey dot quality 3s or something like that

[north yew]

we don't know what the number is, but we know that speed modules have a quality penalty

so generally it's a super bad idea b/c the two modules are interacting in a way that's antagonistic to eachother

[iron root]

that, and applications where youre quality modding tend to not need a ton of speed

[stray marsh]

@iron root what's beam4

[iron root]

bio/energy/astronomics/materials, all t4 😄

[stray marsh]

👁️👁️

[iron root]

theyre space exploration sciences

[daring siren]

BEAM is like STEM but for SE

[rare thistle]

Beam is like @Bean but better under tension

[formal stump]

All I think about beans now is doshs bean power

Lol

[rare thistle]

he's obsessed with me, what can I say

[daring siren]

BEAM is the Erland virtual machine

[rare thistle]

BEAM is also a method for condensing and summarizing spatiotemporal information: Brain Electrical Activity Mapping

[daring siren]

Beam is also "Bean" with an "m" in the end

[terse lagoon]

Mr. Bean? He is there??! 👀

[rare thistle]

Mr. Bean is great. As a former Mr. Bean myself, I would know

[daring siren]

I prefer quality skimming from other processes. For example, when making sciences, use where you can't get prod. Even if it's not for items with useful quality bonuses like .

You get +50% prod on it from the foundry, and 4 modules slots

If you get a high quality you can recycle it and possibly get even better belts

[quaint flower]

why would recycling loops be expensive?
also is the recipe for the same as ?

[daring siren]

So it's a 4x price, but not more than 4x

Recycling loops are expensive because you only get 25% of the items back

[quaint flower]

ohh

dang :\

[daring siren]

If you skim , and recycle those, it's still 1/4, but you're promised quality returns

If it was just 1/4, it would be OK, but when doing recycling loops, you get 25%, and only some of it is quality, so you pay much more as you recycle again

[quaint flower]

that's a alright trade-off for quality

[daring siren]

The idea is to rely on existing high throughput chains, and get the quality items out of them

That said, some is still worth it for 🙂

But if we make , we can recycle the of them and get a lot of

In a roundabout way, every item that doesn't get important quality bonuses, actually helps getting quality ingredients

[quaint flower]

ah.

i think one thing i will try to get quality for is

[north yew]

the is only as useful as your ability to match it with other things of the same quality

it's kinda my beef with skimming off the top strats. it's either not automated or ends up in a big, slowly-accumulating buffer

[half raptor]

The thing about setting up quality crafting at each step is you have to deal with imbalance of qualities and recycling to rebalance. Going prod on every intermediate step and making a quality loop at the end product is probably the way to go.

You can't recycle science packs, but you can recycle loop the ingredients for making science until you get legendary packs if you're so inclined

My previous ideas would be a fun challenge base to see if it's viable, but my first playthrough might just use quality loops for simplicity

Especially for Fulgora where you'll have a lot of common and uncommon which you can recycle down to rare to legendary plates which throws the balance backwards

[north yew]

the more they preview about SA the less I'm in on putting quality mods in non-proddable science recipes ( )

[half raptor]

Remember that quality science is worth 1:2:3:4:6 packs so even hitting uncommon is a 100% increase in science production

[north yew]

• might be kinda handy, but isn't really gonna be usable to upgrade in to other inserter types
• is completely pointless
• is pretty marginally useful since kinda just becomes the thing that makes stone brocks
• is mostly pointless
• doesn't help much b/c you'll still need to make upgrades

oh to be clear I mean skimming the stuff out instead of just using it for packs

[half raptor]

Understandable

I think I'm going to have 2 save files, one for going for quality on every step and one where I prod every step until the final

[north yew]

it's still a little wonky to me to send the quality stuff to packs. just means more machines and your setups are a lot more complicated for a pretty small benefit

• much bigger module cost for those science packs

[half raptor]

You don't have to do quality science at all, it's just a possibility

So if you look at it, putting 4 in an you get a 100% bonus on science production guaranteed

If you use common ingredients in making science with 4 75% will make common worth one pack, 22.5% will be worth 2, 2.25% worth 3... Etc

[north yew]

skimming off , wall, and red ammo might be worth

[half raptor]

So honestly unless you're starting with quality ingredients, you're better off going prod modules for making science

[sick dock]

Man this chat has never died. This mechanic is going to be a forever discussion 🤣

[half raptor]

It's because we all think we are right

😛

[north yew]

yeah absolutely the only reason you'd ever think about doing quality science is b/c the ingredients are not compatible with prod modules

pretty sure even gets you more science than

[half raptor]

Probably

Of course we are still thinking end game strategies, my early game strategy is going to put quality in my bus lines and have a filter splitter put all of the quality into a storage chests and put all the commons on the bus. When I go to craft personal machines, I have chests to draw my early game quality pieces from

What I mean is mk 1 quality modules in my production line and just filter quality to storage and commons to the bus

[north yew]

my main beef with that is that you're not gonna get quality ingredients in the exact ratio of the stuff you want to build

paying a big cost to end up with a bunch of quality ingredients just sitting uselessly in storage chests

whereas if you use or simply no modules you'd get going a lot faster. in your mall is still going to get you a decent amount of stuff, especially considering you can pseudo-recycle by just using normal quality stuff on

like think about how many you slap down before you're even launching rockets in 1.1. even if only 2% of those are quality, that's still more than enough for space platforms

[daring siren]

It's OK to not get quality ingredients in balance naively. That's where you send things to a line vs a line

[north yew]

that's gonna take some very compilicated supply lines and a very large module cost

really slowing down your early game to squeeze out a little bit more quality

in 1.1, it's generally better to make a solution that's just good enough for now and aim at consistent forward progress than it is to try to minmax with the tech you currently have. the main reason being that as you get new techs minmaxing is both easier and pays off more. I don't really expect that to change with SA

you can spend a bunch of time on building these complicated supply lines that take a much larger module cost and much larger resource cost, or you can do the simple way, get to faster, then go back and build better quality infra with the recycler and em plant and

but then wait, we're going to be dumping all of these resources in to recycle loops now, so why not take a trip to to get the foundry and big drill so that we don't burn through 20 copper/iron/coal veins per hour

[kindred crater]

thats why I'm leaning towards volcanus as a first planet (along with getting calcite for nauvis). I agree the mechanics are less "fun" than fulgora, but you're building a solid foundation with what you're getting there

[north yew]

it only seems less fun b/c we don't know about the combat yet

[kindred crater]

yep thats still a big unknown

[north yew]

if planets are designed to be equivalent challenges, then it's almost certain that biters are very dangerous and difficult to work around to bring it up to the challenge level of

[daring siren]

I am not.

Quality loop vs skimming is going to be the next bus vs city blocks 😄

Both are OK and have their advantages and disadvantages

[old lake]

I like that

[daring siren]

There's a lot of merit in quality looping for certain things. Especially in early bootstrapping, and possibly in the very late game which everything is cheap, but in the middle, skimming feels better to me

[north yew]

in terms of forward research progress, skimming is worse than no modules at all. that's my big issue with it.

[daring siren]

Because of the speed reduction?

[north yew]

• module construction costs
• stuff being taken out of the pipeline
• a lot of stuff sitting it buffers waiting for other ingredients of matching quality

[daring siren]

Early-ish game, that buffer is several small chests with quality stuff.

I usually have a chest for personal items like , , and . That chest will now also have versions of things

[north yew]

you still get a decent amount of stuff just from having quality modules in mall assemblers, so is the amount of extra quality really worth paying all of this hefty additional cost and slowing down your early game this much

the need to go tall only really comes in to the picture once you're building platforms and shipping stuff between planets, and by that point you've placed down a lot of normal quality stuff on so your mall assemblers have gone through quite a few cycles

it definitely is starting to feel like 's role is to be the place where you don't need to worry about power and space constraints too much, which means it's the place where you're most okay with using normal quality stuff

[half raptor]

Nauvis is the best place for the Novice

Besf part about early game in everything and putting uncommons and rares in a chest is that if your chest ever fills up and doesn't allow belt to flow all you have to do is take the modules out and let it flow normally and now youve got a steel chest of uncommons and a few rares for making your first modular armor.

Early game can be as janky as it wants to be

In that phase of the game you're not caring about lost raw material and having a rare personal roboport is better than saving 5% raw material

[north yew]

re: modular armor - I'm pretty sure you're paying a greater cost in the supply chain than just having an assembler with quality modules make personal armor and equipment and storing the non-rares until you get the rare outputs you want

assuming you use your best available quality modules in that assembler

with skimming we're halfway sidestepping problem 3 but still running right in to problems 1 and 2 at full force #1215078107334057984 message

[half raptor]

How am I paying a higher cost? At that point of the early game, 95%-98% of assemblers are going to make common and the prod bonus is still small enough that it's not a huge bonus yet. All I'm doing is making a early game bus of common goods but storing the occasional uncommon or rare to the side. If I have enough rare sitting in a chest to straight up make rare armor modules, then it was worth it instead of making 100 armors and hoping for the best

[north yew]

like if the argument is that we don't care about optimizing around raw material efficiency then what we should be optimizing around is logistical complexity and ease of building, right? to make progress in to better tech

in which case just do no modules

[half raptor]

Also cost isn't a big factor at that stage of the game, if you're trying to optimize the early game then you need to work for NASA

[north yew]

#1215078107334057984 message this is the core of my argument in why I don't like skimming

[half raptor]

Until we get recyclers you can't make loops anyway so you build your base for common anyway, but you can set aside the lucky uncommon for more specific crafts

It's not even a complicated thing to build, it's just a filter splitter and a chest or two

[north yew]

if you have quality modules in your machines, your machines are slower. if your machines are slower, you need more machines. more machines, more modules. construction costs are going way up

then space and power and all that jazz are going up too

[half raptor]

If you're going to the early game, your map seed must be sparse

Prod reduces speed too

[north yew]

I'm not even arguing for prod rn

I'm arguing for no modules at all

[half raptor]

Oh

[north yew]

quality intermediates competes with nothing and still loses

[half raptor]

So... with no modules vs some mk 1 common quality modules, you're still making the same number of items (with more machines, yes) you're just storing the few upgrades to the side

The only big investment is the actual making of modules

[north yew]

the biggest cost is how much you're investing in modules alone

but also power and space and belts and inserters and pollution are factors too

once you factor all of that in I'm pretty sure it's cheaper to just make a couple hundred modular armors until you get the lucky roll

[half raptor]

1. You can always build more power
2. You can automate making more machines to accommodate a longer production line easily
3. Space isn't a big ask
4. Pollution is a risk yes, I'll give you this point

[north yew]

idk I guess what I run in to a lot in this server is that people tend to conflate "cheap" and "free" when it comes to things like power and space and machine construction costs

that feels like what's happening here a bit

[half raptor]

I mean coal is somewhat free, though it has a duration. Increase use and you have to tap a new mine sooner

[north yew]

this is the core point though. after all's said and done you're doing a bunch of extra work to arrive at a less efficient solution

[tall sandal]

its because we are long termists at heart and the only thing that matters is the rate of production, capital costs are a trivialrity that goes to zero as time goes to infinity

[north yew]

that mentality makes sense once you've filled out the tech tree for sure

until you've filled out the tech tree, though, it makes a lot less sense

there's an insane rate of return on investing resources in to unlocking new technologies that beats basically anything in the game

[tall sandal]

I agree. I'm explaining the mindset

[north yew]

it's part of why I think it's so neat that SA seems to be bringing space, power, and geometric constraints in to the picture much more. they're interesting challenges to optimize around, but 1.1 doesn't give a whole lot of reason to optimize around them.

[daring siren]

If we start working on early, it negates some of the speed reduction, but I get your point

I agree the main science build would probably not be qualitied, but rather prodded

[tall sandal]

I'm intriged by builds that siphon quality from the the production. they could work

[daring siren]

The main and build for / could be a bit longer, and have in them. A splitter will siphon off items.

Then again, we might want instead

In terms of simplicity, in is probably the easiest, and just work up from there, but it's not cheap, and possibly a bad opportunity cost with

[north yew]

I'm gonna explain why I generally feel like people are overrating the usefulness of squeezing out a bit more quality. This is just an example, and let's not worry about the specifics too much (putting in is generally a bad idea in the early game, this is just to make the point).

• Normal quality items are still useful.
• Quality items are more useful, but in limited contexts. For example, it's generally a bad idea to use speed modules on their own in 1.1 because it's cheaper to just build two machines making the same thing. A similar principle applies here.
• Let's assume for this example that an uncommon is 1.3x as useful as a normal . Rare as 1.6x.
• If we put normal quality in to our and for the for assemblers, we get 1.08x as many inserters for the same resource cost.
• If we instead use for the , we get 1.0066x as much inserter for the same resource cost.
• Measured this way, using prod gets you over 12x the amount of extra inserter for your resource investment vs using quality.

[daring siren]

That tells me: Don't try to . But lets look on other item e.g. or armor/guns or

[north yew]

I just wanted to use a simple recipe for the example

[final nimbus]

quality inserters have a distinct purpose of increasing throughput for really fast machines.

[daring siren]

Quality inserters are irrelevant until late game

Except maybe

[north yew]

my whole point in choosing inserters was to avoid a discussion about what is and isn't worth it to put quality on 😭

[uncut plank]

A few in your 🛰️ assembler might be worth it since it won't slow down enough to need a second one in the early game.

[tall sandal]

I dont really think the idea of siphoning off quality item is necessarily "squeezing out more quality." The idea is just "one day in the future, I want a single set of high quality power armor."

So you toss a few quality modules in some assemblers, and siphon off enough resources to make a single set of power armor. This might be eaisier than quality recycling for a power armor. It might not be!

[daring siren]

I do understand your point, but it changes a bit depending on what you're try to add quality to

Adding does it make it slower, and requires a bit more logistics. Also the cost something compared to nothing.

[north yew]

The concise version of my point is that complex strategies to minmax quality are only worth it if you're only considering best-possible-quality outputs to actually exist, and any case where you're actually using the non-best-quality items makes it clear that it's better to just go prod or no modules.

[tall sandal]

I agree and have argued for a long time that turning your entire factory into quality processing zone is a bad idea. But maybe you just put some quality modules in your , and electric engine assemblers. Then you can make quality power armor for pretty much free!

[north yew]

there's still a huge cost there. lost opportunity for prod, need to have more machines running and more modules built. it'd take quite a few machines running for a long time to get enough for a power armor.

[daring siren]

I think we need to define clearer goals, and see which method achieves that goal

[patent cradle]

The place where I have seen early game quality shine the most (specifically in Colonelwill's stream) has been when chasing bottlenecks in direct insertion builds.

[tall sandal]

When I play, I almost never use prod modules around the time I get power armor because of the costs associated with prod modules (power, machine speed) . Quality modules dont really have a huge cost associated.

[daring siren]

I put in labs and in when I get them

[tall sandal]

Right, but I dont put prods in electric engines and steel

[daring siren]

Same

[tall sandal]

I might put them in advanced circuits, but its not a big deal if I dont

[daring siren]

I do that for my near 1:1:1:1 build
https://storage.googleapis.com/soulburn-mapshot/mapshot/qual/index.html?x=63.7&y=-119.6&z=5.2

[tall sandal]

You can also put the quality module in a single steel producing electric furnace so the impact isnt large. You only need 40 pieces of quality steel to make quality power armor, so it you put modules in one machine and wait long enough, you'll get it

[north yew]

you'd need to make 20,000 steel in that machine on average

[daring siren]

By the time you wait that long, you'll have a larger base so that's not a good wait

[tall sandal]

Are the ods really that bad?

[daring siren]

They are quite bad

[north yew]

is 10% quality strength, so it's a 0.002% chance to get rare if you have two of them in a machine

[daring siren]

With 2 x it's about 2% to get an and 0.2% to get a

[north yew]

☝️ 🤓 2% to get uncommon or better

[daring siren]

Yes

[tall sandal]

2% chance to get an uncommon means I only need to make 2000 steel to get 40 uncommon steel

[daring siren]

We will want

[north yew]

simply "recycle" the normal quality ones by putting them down on

[daring siren]

Ya

One of every 50 will be

[north yew]

and that's with common t1 modules, likely you'll want to upgrade the ones making solar panels ASAP

[tall sandal]

alternatively, if I only have access to , how much steel do I need to use to get an uncommon power armor?

[neat shard]

The tl;dr on quality: If your recipe takes prod mods, use prod mods. If not, use qual mods.

[north yew]

2k

[final nimbus]

#friday-facts message Hmm

[north yew]

ish

[daring siren]

Getting will help too

Also, we'll get about 2 per 🛰️

[tall sandal]

so I'd save 2k steel if I'm able to use quality modules in the electric furnaces

(and electric engine assemblers, and blue circit assemblers)

[final nimbus]

You're not using prod though, so not really "saving"

[north yew]

it also means needing even more furnaces for a thing that already takes a ludicrous amount of furnaces, plus the same effect applying in all of the other ingredients for power armor

[tall sandal]

I'm never going to use a prod module in an electric furnace when I'm on power armor tech

[neat shard]

why not?

[tall sandal]

lol

[neat shard]

prod mods are too good to pass up

[daring siren]

I'm not going to have before late game, except for platforms

[north yew]

are the last thing that get prod modules in 1.1

they have the worst roi

[neat shard]

Huh

TIL

[tall sandal]

Because electric furnaces take an absurd amount of power and prod modules make it even worse

[daring siren]

Well, except for my specific ore furnace

[neat shard]

Even if you don't intend to put prod mods in them, electric furnaces are generally better than steel furnaces, unless the 3x3 size is really that big a dealbreaker

[north yew]

lol that's a hornet's nest

[tall sandal]

They're worse because they cost more power (unless you use eff modules)

[neat shard]

If the increased energy use bugs you, put in effmods and suddenly your electric furnaces use less power than the steel ones

[daring siren]

By better you mean "larger and 2x as energy hungry"?

: "Look how much they need to do to equal our power"

[north yew]

typically I use until I'm able to , but the efficient furnaces are a pretty big upgrade

[tall sandal]

Also, around power armor tech, I'm still happily using my existing smeling lines and have no immedate plans to repalce them

[neat shard]

Admittedly, I'm used to SE where tier 1 effmods are (A) cheap-as-free and (B) better than vanilla tier 2 effmods...

And (C) power is easy to come by in any case

[daring siren]

are late game tech, when you have unlimited power, or for niche cases like space platforms

[tall sandal]

They are good if you want to use eff modules tho

[daring siren]

They are fine deep into

But like , you unlock them much earlier than you need them

I only unlock for the +1 hand size bonuses for normal inserters

[north yew]

costs associated with putting quality modules in and electric engines:

• extra machines needing to run, so extra power, pollution, belts, inserters, etc
• modules needed for all of those machines
• potential opportunity cost of using other modules (although seemingly not applicable in this case)
  when you total all that up, does it really save resources vs just making a bunch of power armors until the lucky roll?

[tall sandal]

I'm not saying you put them in every building. Just a few! Maybe only one! Costs associated with not putting them in those buildings

• 2k
• 2k

[north yew]

especially when you consider that just having a single power armor assembler making a bunch of them makes it very accessible to use the best quality modules available vs using a bunch of cheap ones

so the roll is much higher probability in the final assembler case

instead of doing a low-odds roll thousands and thousands of times, you're doing a higher-odds roll hundreds of times

[neat shard]

Though, granted, each roll is more expensive with the single power armor assembler

[tall sandal]

do we know when we unlock yet?

[final nimbus]

[north yew]

95% certain all T2 modules are on , presumably in a similar place on the tech tree

oh right was confirmed

[tall sandal]

dang thats kinda far into the game

[north yew]

feel like is the transition from early to mid game in SA, like is in 1.1

[tall sandal]

not in terms of % though the game, in terms of hours into the game

[neat shard]

Remember that this expansion is in large part inspired by SE, where beating the game in sub 200 hours is an impressive speedrun

Expect things to take a while

[north yew]

iirc the estimate is double the length of 1.1

[tall sandal]

with remote viewing in the game, do i even CARE about power armor?

i mean, yes. yes I do

[north yew]

also:

• distributing quality rolls across a supply chain to achieve a consistent, predictable EV = boring, stinky
• gambling in a single assembler = exciting, fun, cool

you could get it on the first try!

the big payoff is coming, you've just gotta give it power armor ingredients one more time

[final nimbus]

Theoretically it is most efficient to do every quality roll in whatever assembler has the highest +Quality bonus, meaning the one with 4 x for example.
If you only have one of those, you must use it for everything.
Call it the Slot Machine.

[north yew]

isn't that exactly what they said they would do in the original announcement FFF

Especially for personal armors and equipment, we always ended up having an assembling machine with the best quality modules in the center of the factory dedicated for people trying to gamble for the best possible equipment for their personal use.

feature request: assembler sprite becomes a slot machine if it has significantly better quality modules than anything else on the surface

[neat shard]

Unlikely, but that could be a fun mod

[heavy ember]

Can only insert quality modules to a re skinned assembler that looks like a slot machine

[quaint flower]

i though this would output higher quality stuff. its kinda hard to get

also "But obviously, since items have a chain of steps to produce them, every step has the potential to increase the quality of the intermediate products. With different approaches, and possibly different machines or other ways to improve the productivity of the process (tease of some future FFF content), the cost can be brought lower, but it will always be pretty expensive to get the best stuff."

[kindred crater]

this is where pure volume comes in, and AFK running the game overnight lol

[daring siren]

Just increase production

[daring siren]

With circuits, lamps, and speakers you could make it shine and chime when it gets a high quality item

[neat shard]

Step 1: Just increase production
Step 2: Repeat step 1 until you lag to death
Step 3: Buy better computer
Step 4: Repeat step 2 on new computer
Step 5: Sell stuff to buy even better computer
Step 6: Repeat step 2 on new computer
Step 7: Automate the process of getting and selling stuff to buy progressively better computers
Step 8: Continue to repeat step 2 on new computers
Step 9: Just increase IRL production to sell more stuff to get yet better computers
Step 10: Repeat steps 1-9 until the computer simulating the universe lags to death
...
Step 11: Buy better computer

[wheat sluice]

I had an idea about quality balancing, we could modulate the quality bonus, by using inserters on adjacent beacons to change the number of speed modules affecting the machines. Currently we don't know the quality penalty of speed modules, so I can't tell what it would take, but I think it would be possible.

[kindred crater]

obviously I don't know for certain, but it wouldn't surprise me if all speed modules completely nullify quality regardless of quantity, or do something like -9999% quality

[wheat sluice]

I seriously doubt that that would be the case, but all we do know is:

It is also notable that we created a quality penalty on speed modules, because haste makes waste, and we wanted to reduce the number of places where beacons full of speed modules is the best way to go.

[kindred crater]

to me that sounds like exactly what they're doing here. you could certainly increase production of whatever quality process, but the vertical growth is limited with quality machines, or you can expand horizontally as big as you like

[half raptor]

quality certainly is a very horizontal progression.... but with great reward

I feel like the correct and simplest path is to gamble at the end of the production line and set up a quality loop, ,and it's honestly probably the best way to go. However, a base going for quality in intermediates is a fun path

[final nimbus]

Quality over Quantity or Quantity over Quality?

[half raptor]

the challenge of balancing RNG and all that

My initial thought was to just RNG-cycle excess ingredients, though eventually you'll wind up with more legendary of one ingredient over another and could jam the system and would need to overproduce other stuff to compensate. It would definitely be challenging, but a fun puzzle

[neat shard]

Reward probably isn't worth the effort tho

[half raptor]

but it's fun to design

my first playthrough will probably do productivity on every intermediate and then quality loop finished goods like rational engineer, but my 2nd save file will DEFINITELY try to make my madness work

[kindred crater]

someone will eventually make a blueprint book of stuff that does this with intermediates, and 1/2+ of people using quality will exclusively use those

just like how people use bp books for balancers

[neat shard]

Balancers are self-contained, to make quality work you have to design your whole factory around it

The comparison isn't sensible

[half raptor]

I've been thinking of using the basic city block and circuit controlled recipe setting that changes the recipe quality based on ingredients present by reading the contents of belts, though you still have to deal with the imbalances and that gets tricky if you don't do it righth

[kindred crater]

the bps will be giant and circuit controlled. I guarantee this will happen

[half raptor]

modular designs are my favorite

[kindred crater]

its a cool idea. you'll need to factor in all of the edge/corner cases to the controlling logic

[half raptor]

the thought behind having ciricut conditions set quality of recipe is so that a train can deliver to ingnredients sorted into buffers, circuits control when they unload onto a belt and swaps the recipes accordingly and products go to a - output station and a output station so that you can prod bonus your legendary outputs

[half raptor]

thaht's the thing, I'd have to set up some kind of overflow splitting to a station that will take excesses to a recycler train station, but it feels like the system might accidentally pull too much. I'm going to go into editor extensions eventually and build a test build just to figure out belts and buffer locations

my go to design centers around making a 3 ingredient recipe because if I can balance 3 ingredients, I can do 2 or 4 just as easily

[kindred crater]

yeah. I'm struggling to give my opinion, because it hinges on multiple design choices/production goals

[half raptor]

The problem with trying for quality on each step is that now you have to at minimum have at least 1 and blocks for every ingredient alongside your commons. The issue is that your epic and legendary blocks will run very infrequently and will need a long time for their output train buffer to fill enough to warrant a train coming.

The goal is to make a block that can handle any common to epic input and have the recipe be changed automatically so that we can avoid having a ton of assemblers just sitting idly when you've only got a small handful of epics in your system

my legendary ingredients would be treated separately and go for Prod bonus in their own blocks

[kindred crater]

what is your ultimate goal? to solely output final products? or will you be using to incrementally update your factory?

[half raptor]

incremental upgrades. My for common will also be the chests for a train station with the idea that if my bots replace all of my commons for uncommons, the station will fill and then all of those commons will go to my recycle station and those will feed back to make more once picked up by the trains

that way everything can easily have common modules and once I start getting commons building up, I know I can just upgrade planner the whole base to exchange the commons for uncommon, then rare etc

Of course, I'd manually upgrade the final step assemblers first (or EMPs, whatever optimal design works best)

Once the design is set, my bots would then do all the module magic

[kindred crater]

it'll probably always be EMPs given their prod bonus lol

[half raptor]

right

When I saw Assemblers, it's whatever optimal machinie it is since this is a post-game design idiea

[kindred crater]

understood

[half raptor]

but if I can figure out how to have circuit controlled quality blocks that change their recipes I'll be able to start using the design even into the midgame

[kindred crater]

with this intermediate quality factory, will you focus making one thing at a time?

[half raptor]

so one aspect of the design is that each input has 4 quality buffers and a circuit that does not offload onto a belt unless that quality of the other ingredients is present, so for instance making at my inserters (assuming I have multiple inserters for the same buffer) would be wired to a combinator that says "if > 0 and > 0, then ouput ✅ " which tells it to output the cable onto the belt so that if I don't have the epic ingredients for the other ingredients, then I should just keep my epic cables in theh buffer

[kindred crater]

yeah that makes perfect sense

its a key mechanism for determining you'll actually produce what you're trying to get

[half raptor]

one other thing they said about the circuit controlled assemblers is that a machine will finish the craft it's on if it's midcraft and output any unused ingredients after switching, so I'd have to have an output belt specifically for that which would have to go back to my buffers for my items, which unfortunately would have to be filtered by ingredient then by quality, so a lot of splitter and inserter logic

[kindred crater]

if you don't mind using bots, surely there is some simple way to implement this

[half raptor]

which is fine and dandy but here's the biggest design headache.

Trains will bring in whatever qualities landed in their inventories from a previous production block and there's going to be imbalance. When they unload, I might wind up with a slow but steady buildup of a ingredient which will, given time, make the trains unable to unload

well that is certainly a possibility, though I tend to only use my bots for construction

[kindred crater]

your storage can be buffer chests so if you have circuits controlling requester chests, they won't pull from your main storage but rather the storage/provider chest with the stuff that needs to go back

[half raptor]

true enough, and honestly I'm trying to avoid recycling too many overflowing ingredients as possible

[kindred crater]

you can certainly use belts lol, I'm just trying to save you some pain

[half raptor]

could use a combination that simply refills my "changing recipe" outputs

and then uses belts for crafting

[kindred crater]

if something is overflowing, you must ask whats causing that imbalance (on the other side)- a lack of quality ore, or a lack of production of another intermediate? if its the latter, you can always allocate more machines to establish balance again

quality outputs are of course RNG, but it'll eventually balance itself out

[half raptor]

So here's an issue. As pointed out by practially everyone is going to be an enigma for most quality designs because when making copper has had an extra step for quallity, meaning you'll wind up with a larger amount of quality cable than iron plates. Also that quality cable can combine with green circuits that has had 2 steps now from plate while cable is only 1 step from plate, which imbalances things in a weird way

you can't rely on hard ratios unless you're a spaghetti master chef. You need a robust design that can handle your overflows, either in the form of recycling, or a very careful balance across the whole base

[kindred crater]

you always have the third option of using them for a lower quality recipe, as much as that makes me shudder

[half raptor]

and that's just one thing. Look at it's used for making so in that production line, quality red circuits should logically put their highest quality to near end products, but that's tricky circuitry. Having a design that keeps life simple and balanced is what I'm going for, though I know there's going to be a ton of limiting reagent chasing

[half raptor]

honestly it's not a terrible option

though I do enjoy the idea of recycling and maybe getting some quality ingredients back that are higher than the unused products

[kindred crater]

yeah, it really depends what end quality level of products you're seeking to get (ie, where you're at in incrementally upgrading your factory) and what you want to make, and perhaps what you're going to make in the future

[half raptor]

That gives me a new idea.... still using buffers of higher quality stuff, but using it as overflow for unused quality stuff. Think of this this way. If over the course of many hours you wind up with too much and stations that request the uncommon are always destination full, why not deliver to a factory, like you said and just use the ingredient as an overflow for instances where we just have too many. I even think I know how I'd do it too

it's less recycling and it's using buildups off stuff

building a blueprint for show and tell, give me 3 minutes

[kindred crater]

sure thing

if you're implementing quality in game, you probably want to upgrade as much stuff as possible, using a fairly wide array of intermediates (and with SA and its items/ingredients, it could possibly be a lot more). so looking at it from that perspective, excess stuff for one project could be spent in another not fully rolled out, or put in recycler loops. I probably sound all over the place here, just because there are so many angles to this

[half raptor]

the thought of using excess

[kindred crater]

what you're going to make next is indeed another piece of this puzzle

[half raptor]

that's an idea that I'd have to noodle on

damn my game is loading slow, might just have to use a website to draw it real fast

[kindred crater]

there's always going to be varying levels of "waste", you may have to do the math to see which strategies minimize that

waste = recycling, or using higher quality for level quality recipes. well I take that back, you can eliminate that waste with balancing like you said a bit earlier.

[half raptor]

okay so I'll need to restart my pc to load my game, but I made this on the blueprint editor site, so it's very basic proof of concept

color coded to the best of my ability

the first 6 inserters grab commons that go to the far left station

but any quality excesses go to the 2nd station on the right so that it lines up with these chests for varying qualities

and just sacrifice them to be used as lower grade ingredients if my input buffers are full elsewhere

[kindred crater]

perhaps crafting 2 things you want at the same time, one is iron heavy for instance, one is copper heavy, so in a way you have more built in stability so things don't reach gridlock. or multiple items if their amount of ingredients is higher. I'm just thinking aloud, hopefully you get some value out of it. reading what you typed now.

thats a very clever design, I see what you're doing there

[half raptor]

alternantively I just have a regular station design and just train logic "if =/= 0, go to station. Interrupt if Destination Full AND time elapsed =10 seconds (because a station could open up) then go to station and unload

though one thought is that each station would have filters set on inserters in case a recycler train has like 10 different items on it, which is why having a longer station like above would give me buffers specifically for each quality to be unloaded

[kindred crater]

I understand, was just thinking about your recycler train statement

so recycling for all items would be done in one location, and some station in which all items get unloaded

[half raptor]

a very generic, repeatable recycle station

that way I don't nened to set up recyclers in every block, just an output station that hopefully would not need to get used. Then a generic train would pick up all the ouputs and delivery them via interrupts

[wispy dragon]

Blueprint String

Blueprint 1

[kindred crater]

that makes sense

[half raptor]

I can also set priorities for quality overflow delivery stations so that excess rares will deliver to uncommons first before commons, the thought is that my quality blocks should be so starved that they shouldn't need to overflow often but if they do, they need an outlet to keep production moving. I'd rather sacrifice a train load of common green circuits just so I can produce another train load and maybe 25% of that production becomes quality and the ingredients from the recycling can also be quallity

But the idea of using them as lower quality ingredients first is a good way to still use them for crafting long before I even think about recycling then and frankly if all of my epic input buffers are full, then damn my base is doing good

Legendary can just buffer and wait their turns at the depots for all I care

The only problem with the design proposed above is that it can't be used in combination with my idea of a block that changes recipes based on what ingredients are available

Oh God wait.... I'm dumb.... My unloading with sideways splitters allows for an input.... Like if my assemblers change recipes and spit out various ingredients....

I could just plug it back in to the original input....

[kindred crater]

yep there you go lol

[half raptor]

But naw that doesn't work either since you can have a belt full of epic of one ingredient and rare of another.... I got excited too quickly

I think I'm doomed to have a block for every quality of every ingredient

[kindred crater]

I wish I had more useful things to say, but for me at least with quality I need to see the full system to understand what its doing and then see how things can be modified.

[half raptor]

Will tinker with the idea though

[kindred crater]

maybe you can have some kind of circuit for detecting if something shouldn't be in the network for that given block, and that by default moves it "somewhere" to be removed

[half raptor]

I mean a massive bot base would solve all logistics, but I'm not a fan of bot swarms

ILikeTrains.meme

[kindred crater]

its certainly ambitious, but it should be quite efficient w/ material

[half raptor]

Every time we get a new train feature or circuit feature my strategy for changes

[kindred crater]

can't believe I've been talking about this for over an hr lol

[half raptor]

I can't believe I've been talking about this for 6 months

Ive never talked so much about the same game

[kindred crater]

quality is definitely one of those mechanics where someone could probably write a 1-200 page book on strategy and still not cover it all

[half raptor]

Yeah then someone writes a blueprint that solves it and it becomes the new staple blueprint in everyone's base

[kindred crater]

yep lol. just put what you want in a constant combinator connected at the bottom corner of the BP, and let it chug away

[half raptor]

Even easier with parameterized blueprints. Go to your 2 ingredient +1 fluid blueprint, select and it just sets everything

[kindred crater]

ah yeah, forgot about those things

[half raptor]

I did completely forget you can use higher grade ingredients with lower grade if you had to. Great way to use the item instead of throwing it in the woodchipper

[kindred crater]

especially if when in the early-mid stages of incrementally updating, you could use that lower quality intermediate

but yeah, a lot of moving pieces in such a universal design. it would be seriously impressive to see someone pull it off though and have it run in an intelligent way

gotta run for now, but thanks for giving me a lot to think about haha

[half raptor]

Most intelligent designs in this game starts with a bullet list of what you want it to do, a list of potential problems and workshop ping how to overcome them. That's why I like the idea of a repeatable recycler station because if my whole base only uses one recycle station, that's wonderful, but my base might want 4 or 5 for various reasons

[kindred crater]

definitely. there are a lot of parallels with software development and the importance of planning before execution. and like you said, I'm sure there will be more things revealed (such as that more controversial mechanic mentioned) which will once again have us going back to the whiteboard and starting over lol

cya

[north yew]

what about using circuit controls to make it so that inserters will only output in matched swings
eg in the section inserters unload 3x for every .

use the any quality recipe on the assemblers

[half raptor]

Good idea too

[north yew]

I'm pretty sure you would rather pair high to low quality in the case of overload than recycle the high in most cases

[half raptor]

As far as modular designs go (assuming city train block) the more smarts and parts you put into the block the less room you have for actual production, so it has to be efficient and tidy

[half raptor]

I'm on board with this idea, I think that i would only need to set a recycler station condition if for instance the station destination full and :cable stations are destination full. Even if this edge case NEVER happens, I'd still like to build with this in mind so that even if things really get bad, I've got a last resort

Reducing quality of ingredients is definitely preferred over recycling 9 times out of 10

That is also the reason behind having a separate station for recycling as opposed to local recycling because I can have just one line of recyclers handle most of my base's needs for recycling if I build this right, and also I can have my best quality modules on the recycler line until my whole base if legen (wait for it) dairy

[kindred crater]

@half raptor I can't stick around but leaving you with a final puzzle piece that I didn't think of earlier- productivity research. the more you research, there will eventually be some objective/subjective threshold where it becomes "worth it" to recycle loop, if not for the end items themselves, a reduced cost way to get the higher quality intermediates of it. so your strategy for processing can change as research for given items progress.

[half raptor]

and for that, I can change my train schedule instructions accordingly, though having the station would give me the ability to change my tactic based on current state

thanks for the discussion, it was helpful

[half raptor]

The more I play with this lowering quality idea in editor extension the more I realize this is almost a pre-Fulgora form of recycling where you're not actively losing material

I just set the last chest to unload non common stuff only, reverse the direction of the splitter and then prioritize it so that it unloads first. Now if I wind up with an excess of quality material, I can just tell it to go down the list of qualities until it hits common and then if all quality downgrades are destinantion full, then and only then would it go to the recycler

[iron root]

this makes me think...
Can you limit a 's hand size to guarantee a certain number of items are swung to the destination

[half raptor]

yes

you can do that in 1.1 right now

limiting stack size in the inserter

[iron root]

nah with it can grab less though

[half raptor]

well that's true and a stacker inserter will only swing if it's full and a bulk inserter will swing once its had a pause to wait for more material, so probably not what you're thinking

[iron root]

which is why I'm asking if I can limit to make it swing with a very specific # of items

no more, no less

[half raptor]

with our current information, i'd say no, but that's a good question for the devs, which I don't see any in this channel so you may need to ask in the friday facts or space age channel

[iron root]

suppose ill bounce it to the fff channel

[half raptor]

so I have a feeling devs get a little ping when someone says their name, so perhaps ask them directly by name and they might answer

[iron root]

itll happen sooner or later

[jaunty citrus]

I'm not sure who else has their name highlighted in the highlight bot, but I know does.
I also have their name highlighted. I assume a lot of people do at this point.

[sacred totem]

I have a specific phrase highlighted

It has yet to appear

[neat shard]

Hm... Asriel Dreemurr?

[sacred totem]

Good idea. I’ll add that one

…if I can remember how

Yup, remembered how

[jaunty citrus]

Should I highlight statements like "I don't like quality" or "I think quality is bad" so I can always be in the chat to argue with whoever said it?

[sacred totem]

No, you have better things to do with your time

[jaunty citrus]

Hmmm, fair point. I'll need that time to sleep so I have less awake time between friday facts

[daring siren]

😄

[half raptor]

If you come to this channel to say quality is bad, that is fine but you better argue your claim and prove it to us

I already know some people who will stubbornly play Space Age with no quality and that's their choice

[little orchid]

I've always considered modules and especially beacons to be something that many players tend to not approach because of how "weird" they are compared to how the game has been up to that point

And they can play just fine without them

Quality will be no different

[half raptor]

In 1.1 i play with no uranium

SA is going to force me to accept it though

[jaunty citrus]

ditto

[kindred crater]

and a lot of it

[old lake]

1.1 is completely doable without uranium and a lot of people probably skip it. It’s an endgame resource used as a more powerful version of three things that are all accomplished in some less tedious way by something else, and not necessary for 1.1 science. Don’t get me wrong it’s super fun and useful, but if you look at uranium and say “nah too many steps” that’s fine by me. But you can’t really do the same thing with quality…

[charred gyro]

You can't? Or is quality just much more useful, so there's more incentive to try it?

[north yew]

they said in the announcement that it's technically possible to beat the game without ever making a quality module

just gonna be tough

[final nimbus]

Supposedly it's balanced around not using quality.

And quality itself is expensive/rare enough that using it can't be obviously better.

[daring siren]

I only invest in uranium if I expect a longer game. It requires a large upfront cost, and produces way more than I need before endgame.

[chrome mauve]

yeah it is good that uranium is getting some love

it was sad that you had one mine that would last houndreds of hours without drying up

as for quality I think of starting it from minnig ore and then go upward

[final nimbus]

I always thought uranium was too abundant of a resource for something you need very little of.
And that's if you choose to mess with it at all.

[kindred crater]

agreed. I wonder if there are other unknown uses for uranium in SA we don't know about yet as well, beyond science and reactors on platforms

[tall sandal]

I suspect that light use of quality will be optimal for a regular playthough of SA, and zero use of quality will be somewhat close to optimal.

[half raptor]

Honestly I find solar to be my favorite and biters never got bad enough to justify uranium bullets

[gritty umbra]

My opinion on quality is going to heavily depend on the tools we have for managing inventory, construction, and so on. For example, how hard is it to put down a blueprint that uses quality buildings if available, normal if not, and doesn't use Q5 at all?

[daring siren]

We currently don't know if such systems exist. It could be useful for progressive upgrading of modules too.

[gritty umbra]

Since this system introduces five variants of every building, it could end up very clunky.

I'm hopeful that the Factorio devs will supply the tools to manage that well, but it's currently the main reason I'm reserving judgement on quality.

[tall sandal]

I cant think of a lot of use cases where I'd be okay with an unknown number of buildings being an upgraded quality

[wheat trellis]

Where can I read about quality? I don't know what it refers to.

[tall sandal]

https://www.factorio.com/blog/post/fff-375

[wheat trellis]

wow

[gritty umbra]

I can see situations where I don't care if a building is Q1-Q3, but do care that it doesn't steal away Q4 or Q5 that I want to save for specific builds.

[tall sandal]

I'm sure they exist, but none really come to mind

Can you share some?

[gritty umbra]

General purpose builds vs space platforms, for one example.

[tall sandal]

I dont really know what you mean, could you be more specific? Here's one specific example where I wouldnt be okay with a few buildings being upgraded quality -- my smeling line. I size smelting lines to fill a belt, so randomly adding a rare smelter to my smelting line wouldnt do anything.

[tall sandal]

Let us know what you think when you finish reading!

[gritty umbra]

More generally, any situation where the number of buildings I can place is limited is one that I want to save the highest quality buildings for.

But since the quality system will spit out a lot of Q1, Q2, and even Q3 buildings for every Q5, I don't really have to care what quality it uses for an ordinary build, as long as enough buildings get built.

[tall sandal]

Oh so you are saying that q2 is trash to you, so you'd rather see your smelting line get built faster by using a q2 smelter rather than waiting for a q1 smelter to be finished?

[gritty umbra]

The mechanics seem to lean that way, especially as you get higher tier and higher quality modules to play with.

[north yew]

as far as blueprints and such goes, it works like belts and inserters for 1.1

upgrade planner interacts with quality

so you can upgrade all the assemblers in a bp to a different quality, for example

[gritty umbra]

This wouldn't be about upgrading.

[tall sandal]

That's really interesting because I think I'd want to save my q2 assemblers.

[gritty umbra]

This would be about saying "This building in this blueprint can be Q1-Q3, use whatever's available".

[north yew]

why would you want to do that?

[gritty umbra]

As noted above, the large amount of lower quality buildings for each higher quality building.

[north yew]

presumably you're building things to get something in the ballpark of a specific level of consumption or output

[tall sandal]

Scroll up, they explained why here. Basically they're okay burning medium quality building if it means their stuff gets built faster.

[north yew]

the distinction in how useful an item is and its cost/rarity is gonna make situations where you simply don't care what quality something is very rare imho

[gritty umbra]

In this situation, I'd plan around Q1 and not really care if Q2 or Q3 gets used if that's what's available, in a situation where I'm making the highest quality buildings for special purpose builds.

[north yew]

like it'd be the same thing of having a bp option to just use interchangably

rare you'd ever want to do that

if the setup is designed to work with normal quality, then just build normal quality buildings with it and recycle loop the uncommons and rares. that's a better use of the resources anyway

[gritty umbra]

Very debatable.

[north yew]

I really don't think it is

normal quality buildings are going to be substantially cheaper and more common than even just uncommons

uncommons are valuable even if you don't explicitly need uncommons now b/c they have a much better chance of spitting out an upgrade when recycled

like your odds of a legendary output when recycling or crafting from uncommon go up by a factor of 10

[gritty umbra]

That very much depends on how you're making them. If you're focused on recycling them, then yeah, it's going to get expensive. But I see quality more as a function of binning the components at each step, letting the quality improvements compound.

[north yew]

oh boy

[daring siren]

It's always fun seeing people reading this for the first time 😄
You should catch up on all FFFs since 373 up the current, 402

[north yew]

#1215078107334057984 message

tl;dr you don't want to do quality on intermediate production

even just as recycle fodder, uncommons are worth 10 commons. rares are worth 10 uncommons. epics are worth 10 rares. you don't want to throw around quality willy-nilly

[tall sandal]

but I want you to do quality for intermediate production

[gritty umbra]

I really don't see the issue. Beacons are irrelevant, since I never use them. Ratio? Quality is for infrastructure, so I'm not concerned about ratios. Complexity? Yeah, it's going to be five belts for each input and five for each output, with splitters to help with the binning. But that's the sort of thing blueprints exist for.

[north yew]

if you roll for quality on when making , the ratio of high quality to low is different between the and . the roll for quality on cables was entirely wasted

that problem repeats itself over and over and over and over throughout the supply line

[gritty umbra]

"Entirely" is false.

[north yew]

what benefit do you get from rolling quality on the cables

[gritty umbra]

So that I can combine it with the high quality iron plates, of course.

[tall sandal]

this is okay because you can just put in your assembers instead

which will solve the ratio problem for circuits

[north yew]

if you follow that through on the whole supply chain what you end up doing is just putting prod in everything

[tall sandal]

hmmm I dont think so

[north yew]

like you only get one effective quality step if you roll quality on t1 modules, t2 modules, and t3 modules

because of the way those recipes all feed back in to eachother

[north yew]

how are you getting more high quality out of your furnaces than high quality

[unborn flax]

just make one probability distribution match the other

[north yew]

how

[unborn flax]

send iron through quality one more time

easy

[north yew]

lol

[tall sandal]

you take your quality (you will have a lot of these) and combine them with your quality to make quality . Then you wont have a lot of quality left over which fixes your ratio problem for

[north yew]

it's not gonna just magically work out to the numbers you want it to

you'll have excesses of some items of some qualities and shortages of other items of other qualities all the time

[gritty umbra]

If I end up with an overabundance of some component, I'd simply cascade it into the lower tier.

[north yew]

which would be fine if productivity modules didn't exist

[unborn flax]

you can make the probability distributions match exactly in this case

[tall sandal]

but im not normally using productivily modules

[north yew]

why not?

they're quite good

[tall sandal]

because they make my machines slower and cost more power

[north yew]

quality modules also do that

[final nimbus]

[tall sandal]

and also they are costly to manufacture

[tall sandal]

much much less though

[gritty umbra]

And quality modules aren't expensive?

[north yew]

consider: it's not just the specific machine the prod modules are in. it's everything behind them too. even just making means you need only 93% of the drills running, 93% of the trains moving, 93% of the furnaces smelting for the same green chip output

[gritty umbra]

So you're advocating for pure productivity up until the final step?

[north yew]

and that effect stacks up multiplicatively

[north yew]

absolutely

or no modules

if you're in to that sort of thing

[gritty umbra]

I'm curious if you've run the numbers on the input resources for a given quality output.

[north yew]

I have

[gritty umbra]

Because the impression I've been given is that only using quality at the last step and recycling until you get the highest quality available is the least efficient way to get there.

[unborn flax]

pure productivity until the final step is definitely a good strat, though the final step will typically be non-productivity-allowing, so numerically suboptimal if you care about that

[north yew]

well yeah you absolutely put quality modules in the final step that's not up for debate

[unborn flax]

i mean that pure quality modules isn't the best way to do quality recycling loops

[north yew]

the people who have given you that impression are only considering the math in terms of a single step and not a whole supply chain

[unborn flax]

obviously you load a machine with q modules if you have no other choice

[gritty umbra]

And the proportion of quality components also compounds with each step that you use quality in.

[tall sandal]

i know prod modules are good. but i dont ususally use them until super late game

[north yew]

then for the same reasons you should avoid using quality modules in your intermediates

[tall sandal]

i am, but for seperate reasons

[north yew]

this is not the case. the ratio problem prevents it from being the case.

[daring siren]

For recycling, 2p2q is best with q5t3

[gritty umbra]

It reduces the amount of compounding but does not prevent it.

[tall sandal]

you can in theory also solve the ratio problem by blending prod modules and quality modules at the right ratio

[unborn flax]

do you think the below distributions are different?

1. copper plates --(qual)--> copper cable
2. iron plates --(prod)--> iron gears --(qual)-> iron plates

[north yew]

in the example I gave earlier, t3 modules, you're only getting one effective quality roll out of putting quality modules in the assemblers for like eight different products

[north yew]

correcting it this way is exactly as wasteful as recycle looping at the end

[unborn flax]

so you think they have the same distro?

[north yew]

no, because recycle loses 75%

[unborn flax]

they have the distribution of relative qualities

you can throw away 99% of the items if you want and they'll still have the same distribution

[north yew]

oh yeah if you don't put quality modules in the recycler I guess

but now we're being more wasteful than recycle loop at the end

[gritty umbra]

May I see your numbers?

[unborn flax]

you do put quality modules in the recycler - the same Q power as in the copper cable recipe

[gritty umbra]

And for the sake of discussion I'm willing to assume T3 quality modules with +2.5% quality each.

[north yew]

then now the iron plates coming out of the recycler are a higher quality than the copper cables, so now you have the same problem in reverse

[unborn flax]

no they're exactly the same

[north yew]

okay we assume iron plates and copper plates are coming out at the same ratio
copper plates to copper cables is one additional quality roll
iron plates to gears is one additional quality roll
gears back to iron plates is another additional quality roll
the iron plates have two rolls on them now, the cables have one

also you've just quadrupled the cost of iron plates

[final nimbus]

Throw into recyclers
That's one roll

[unborn flax]

i said "--(prod)-->" in my plates to gears line - if this wasn't clear, it means you use productivity only

[north yew]

oh

sure in that case you have balanced the quality ratio for one specific item by increasing the cost of one of those items by 4x

now do it again for the whole supply chain

you'll have to do the same thing with when you're making because the red chips take

then you'll have to do it again when you're making because the and ratios are different

[unborn flax]

but you no longer think you can't match the distributions, right?

[north yew]

sure it's possible but it's honestly kinda silly to do it this way

significantly more wasteful and more logistically complicated and more costly to build than just recycle looping at the end

[unborn flax]

well yes but equally it's kinda silly to theorise over quality

this is one of the reasons quality is a really good addition

the 'prod everything to last step' solution is great but you probably have a numerically suboptimal quality loop at the end (no prod modules allowed)

whereas doing it lower down gives probably better ratios, at the cost of massively nerdsniping yourself

[tall sandal]

i think its a solvable problem

[north yew]

this way of balancing each specific ingredient is numerically worse than recycle looping at the end

[tall sandal]

hmmm I dont see how thats possible

[unborn flax]

there are other ways to get distros to balance, or at least balance more closely

[north yew]

even in the best case scenario, you're still going to need to recycle loop in at the end if you want more than a trickle of high quality items

[unborn flax]

what do we know so far about quality and science packs?

[final nimbus]

Quality you can scrape from the factory also solving the ratio problem there.
That is, the assemblers use but all the quality gets sent to instead.
The real problem will be the

[north yew]

it won't balance out like you want it to

[unborn flax]

please stop saying that

[tall sandal]

they have more science per science

[unborn flax]

do we know the exact % per tier?

[gritty umbra]

Probably the same as other quality bonuses?

[tall sandal]

like 100% or something. legendary has 600% science iirc

[unborn flax]

i think a good problem for quality people to think about is red science with the 1.1 recipe (copper plates + gears)

because it's really simple, it's technically a bit unbalanaced and (unlike say electric poles) we need them in bulk in the endgame

[gritty umbra]

Also a situation where recycling is a horrendous solution.

The best quality bonus is +150%, and that in no way can offset the the 75% loss with recycling.

[unborn flax]

we might have to pretend the molten foundry recipe doesn't exist if we don't have all the information we need

[north yew]

quality science is a very bad idea generally. the only time you would consider it is in science packs where the ingredients aren't proddable, like

[final nimbus]

Quality science has always been a bad idea unless you're trying to void excess quality bits and bobs.

[north yew]

you get way more science per science by just using prod on the ingredients

[unborn flax]

that's assuming we have 30% per quality tier right?

[north yew]

it's a bigger bonus than that and even then it's horrible compared to prod

[gritty umbra]

Not really, no.

[unborn flax]

the question I have is "what % science improvement (per tier) would be required in order to maximally nerdsnipe us?"

[unborn flax]

isn't that precisely +30% per quality tier...?

[north yew]

like an uncommon science pack would have to be 30-40x as much science as a normal quality one to even think about putting quality modules in

[tall sandal]

I think for science its

• Normal: 100% science
• Uncommon: 200% science
• Rare: 300% science
• Epic: 400% science
• Legendary: 600% science

[gritty umbra]

No, look at Q4 to Q5.

[unborn flax]

I understand Q5 to be two tiers above Q4

[tall sandal]

heh

[gritty umbra]

But there's no tier that exists between them.

[final nimbus]

It skips a tier so legendary is more appealing.

[unborn flax]

was this from a dev post or FFF?

[tall sandal]

let me go ctrl F 600% in the FFF channel rq

#friday-facts message
#friday-facts message

boskid — 09/08/2023 7:35 AM
if you somehow get a legendary science pack, it will contain 6x as much science in it as a normal one

boskid — 09/08/2023 7:36 AM
No, that implies every quality level is +100% for science

[unborn flax]

for these numbers I could believe the optimal strategy isn't just prodding everything

[north yew]

it is

you don't want to do quality science

[unborn flax]

you may well be right but i have yet to run the numbers myself

[tall sandal]

lmao boskid

not my fault you turned that ping on

lets say you want 50 red science and 50 rare yellow belts

[unborn flax]

i could believe e.g. you would need at least +x% prod in certain endgame recipes for it to work

[north yew]

is 10% guaranteed, is 2.5% chance to double

like it's not even close

[tall sandal]

lets say you have access to .... t1 modules of no quality

[unborn flax]

in fact here's a counterexample: with endgame prod bonuses maxed out at +300% for every recipe, there is no reason not to produce legendary and legendary

[north yew]

then it's still 4% guaranteed vs 1% chance to double

there isn't gonna be a prod research for every item

[tall sandal]

and of course lets say you have access to a recylcer

and because why not, you have access to electric furnaces

[north yew]

also probably none of us in this discussion are going to play on a single save long enough to hit 300% prod on any of the productivity researches

it takes weeks of your factory running even at insane SPM levels, like 50k

[kindred crater]

I'm going to try in all seriousness

the plan is to run AFK for months

[unborn flax]

that's not the point - say we agree that after running your factory for a year, quality is a good idea. then there's a point before that at which it first becomes useful and i want to know where that is

[north yew]

you're assuming there's an infinite prod research for and , which there likely isn't

[tall sandal]

Option 1: full prod in miners, furnaces, assemblers (except yellow belt assemblers), quality recycler loop to get your rare yellow belts
Option 2: full quality in miners, furnaces, full prod in gear assembler and red science assembler, quality recycler loop to get rare yellow belts

[north yew]

why are we recycling for rare red science

[tall sandal]

we are recycling for rare yellow belts

[north yew]

why are we recycling for rare yellow belts

[tall sandal]

both setups want to produce 50 rare yellow belts and the equavalent of 50 red science packs

[tall sandal]

because its a stand for a finished product you might want to get quality of

[north yew]

it's gonna be at the point where prod modules are putting you over the cap so the prod is completely wasted

you can't recycle science packs btw

you can, but it'll be like : spits out the same thing 25% of the time

and spits out nothing otherwise

[tall sandal]

did anyone suggest recycling science packs?

[north yew]

here

[tall sandal]

I dont think I suggested recycling science packs there

[north yew]

yeah mb the example got broken up by other bits of chat

[charred gyro]

quality science is a thing?

[tall sandal]

yes, see

[tall sandal]

the question is: which factory will use less ore to produce 50 rare yellow belts and the equavalent of 50 red science packs?

[north yew]

in this specific example, the second option is cheaper, but only because both of these things are true:

• yellow belts have a very short chain and are bottlenecked by the ratio problem very little
• you're enforcing a strategy of recycle looping until best possible quality

#1215078107334057984 message

[tall sandal]

I picked those recipes simply because they were super early game and easy to reason about. I'm not sure what you meant by point 2.

[north yew]

recycle looping for only best possible quality is likely to be a generally bad strategy and huge noob trap

if you value not-best-possible-quality at all then prod blows quality out of the water, basically

[tall sandal]

so in your opinion, the goal of 50 rare yellow belts is a bad one?

[north yew]

yes, whatever the belts are substituting for

[tall sandal]

I mean obviously rare yellow belts are silly at all

[north yew]

additionally, if you applied this strategy to a longer, more complicated chain, you'd realize how few quality modules you actually get to use and how much prod you'd be using instead

for example: roboframes. has a very short chain, so all of the intermediate steps between furnaces and roboframes are gonna be prodded

[tall sandal]

what would be a better goal? 50 normal yellow belts and 50 rares? or something else?

[gritty umbra]

Try inserters, since those benefit from quality on their own.

[north yew]

simply measure it like I did in that message I linked

I guarantee you'll be getting a much larger amount of yellow belt from the prod approach

[tall sandal]

steel is 3 steps: miner -> smelt -> smelt
circuit is 3 steps: miner -> smelt -> circuit
electric engine is 4 steps: miner -> smelt -> engine -> electric engine

[gritty umbra]

While any given end product is effectively limited by the chain with the fewest steps in it, I'm not going to be building an optimal factory for every single end product, so I'm still going to want quality in other steps.

[north yew]

batteries are a short chain too

[tall sandal]

batteries are 3 steps: miner -> smelt -> battery

[north yew]

and you'd have to have two segregated areas where get manufactured: some that go to roboframes and some that go to electric engines

b/c you don't want to roll quality on the ones going to electric engines

[tall sandal]

it seems like the optimal setup would be something like:

• quality in miners, smelting, batteries, engine, robot frames
• prod in gear, copper wire, electric engine

[gritty umbra]

Okay, that one is true. Fluids don't get quality. But I'm not going to be making electronic circuits for electric motors separately from other circuits.

[north yew]

this way of solving the ratio problem does work, but it massively exacerbates my other two main problems: scale and logistical complexity

[final nimbus]

Sounds fun

[north yew]

like I've known about it since I started arguing about the ratio problem months and months ago, but I just generally try not to spread it b/c I think it's a bad idea

it just has an exponential explosion in the complexity, size, and construction costs in your base b/c you're having to make every product with different numbers of quality modules in different amounts

then making sure that the lines where you're making green chips with one quality module stay segregated from the lines where you're making green chips with two quality modules

[gritty umbra]

That complexity only exists if you choose to pursue that degree of complexity.

[tall sandal]

it just has an exponential explosion in the complexity, size, and construction costs in your base
I agree with this. Its why I wont be doing quality in intermediates (for my first playthough)

I'm definetly doing a playthough where every building is required to have at least one quality module though

[gritty umbra]

For me, I'm just going to be making a block for each step with quality. If that results in suboptimal ratios for any given end product, so it goes.